Train Location System Using Kalman Filter and Milepost Resets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle location determination systems for trains face challenges in providing accurate and robust real-time location information due to errors and disturbances in GPS and speed measurements, particularly in railroad environments, where transient errors and intermittent lapses are common.
Innovation Solution
A vehicle location determination system that utilizes GPS information and vehicle speed data, combined with a Kalman Filter to estimate real-time location, compensating for errors and disturbances, and employing critical mileposts for reset and diagnostic purposes to ensure accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If GPS and speed measurement methods are used to determine vehicle location, then real-time location information can be obtained, but errors and disturbances accumulate leading to reduced accuracy
Solution Approach 1:
The system employs a feedback mechanism where the estimated position is continuously compared with actual GPS positions when available. The Kalman filter uses this feedback to adjust its estimates and reduce errors. The system also provides feedback through milepost comparisons to detect and correct cumulative position errors, ensuring long-term accuracy despite measurement disturbances.
Solution Approach 2:
The Kalman filter acts as an intermediary that processes and fuses GPS measurements with speed sensor data. It mediates between the two measurement sources, optimally combining them to produce a more accurate position estimate than either source could provide alone, while filtering out noise and disturbances from both inputs.
2Adaptability or versatility
If dead reckoning method is used to compute position by integrating speed measurements, then location can be determined without GPS, but cumulative distance errors increase over time
Solution Approach 1:
The system performs preliminary actions by establishing a known reference position at mileposts before continuing with dead reckoning. This periodic resetting prevents cumulative error buildup by re-synchronizing the integrated position with actual GPS or known track positions at predetermined intervals, maintaining accuracy over extended operation.
Solution Approach 2:
The system implements periodic correction at mileposts where the estimated position is compared with known track positions. This periodic action resets cumulative errors that accumulate during dead reckoning intervals, enabling the system to maintain precision over long distances while operating in GPS-denied environments between mileposts.
3Measurement precision
If Kalman filter is used with GPS and acceleration measurements, then position tracking can be improved, but certain errors and disturbances such as cumulative distance errors are not addressed
Solution Approach 1:
The system transitions from a standard three-dimensional Kalman filter to a four-dimensional formulation by adding the cumulative distance error as a separate state variable. This dimensional expansion allows the filter to explicitly model and correct for cumulative errors that the traditional three-dimensional filter cannot address, improving robustness while maintaining precision.
4Measurement precision
If multiple sensors are employed for location determination, then accuracy can be maintained, but the system becomes vulnerable to transient errors and intermittent lapses
Solution Approach 1:
The system implements beforehand cushioning by using the Kalman filter's predictive capability to maintain accurate position estimates even when GPS or speed sensors temporarily fail. The filter's state prediction continues to provide reliable position information during sensor lapses, cushioning against the impact of transient errors and intermittent failures.
Data Source
AI summary
A location determination system includes one or more processors configured to determine a location of a vehicle system based on output from a location determination system onboard the vehicle system. The one or more processors also are configured to determine whether the location of the vehicle system as determined by the location determination system is accurate based on one or more of performance of the vehicle system, communication between two or more vehicles in the vehicle system, one or more other locations determined by one or more other location determination systems onboard the vehicle system, and/or operation of a brake system of the vehicle system.